Donald O. Hebb conceived of the cell assembly - a small network whose neurons fire in stereotyped sequences - to represent a memory trace. These cell assemblies are believed to underlie persistent activity patterns observed in many experimental preparations. However, it is unknown how these assemblies develop ex nihilo from a population of developing neurons, and how an assembly can retain stability once it emerges. The first of these questions bears directly on how memory traces can emerge in the brain. The second question is of direct relevance to epilepsy, a pathology characterized by unstable, hyperexcitable neuronal networks. Rat hippocampal culture provides a model system for studying these questions. In addition to being a highly tractable system, it supports the spontaneous development of reverberatory circuits, a hallmark of cell assemblies. Using this system in combination with perforated patch clamp recordings, I will examine the reciprocal relationship between activity-dependent synaptic plasticity and Hebbian cell assemblies. [unreadable] [unreadable]